The Languages of Landscape.
MARK ROSKILL is a distinguished art historian and this book, despite a cover blurb, which badly exaggerates its theoretical ambitions, is essentially an exercise in art history. It discusses, in broadly chronological order--from the Akrotiri frescoes of 1500 BCE to twentieth-century American `abstract landscapes'--some main episodes and themes in the story of Western landscape art. Those episodes include the shift from a primarily narrative use of background landscape scenes to artists' concern for nature `for itself, and from paintings' provision of information about the natural world to their `evocation' of it. Some main themes are the relation between painters and their clientele--eighteenth-century landowners, nineteenth-century tourists, and so on--and that between styles of landscape and literary genres or fashions. Roskill's history does not attempt to be a comprehensive one. In comparison with, say, Simon Schama's Landscape and Memory, he provides relatively little either on art's response to shifting conceptions of the natural world or on the national, cultural resonances of different kinds of natural scenery. Moreover, there are many important landscape painters and movements that receive little or no mention--nineteenth-century Scandinavian and twentieth-century British ones, for example.
Considered as interpretative art history, the book is something of a curate's egg, the `good parts' often closely juxtaposed with the `bad parts'. Thus, on pp. 202ff. Roskill distinguishes two trends in the interwar art of our century., the `landscape of presence' and the `landscape of tourism'. The former is strangely labelled, obscurely defined, and not provided with any accompanying illustrations. The latter is appropriately labelled, crisply characterized, and helpfully illustrated. (Although the book contains some fine illustrations, these are insufficient in number. Not a single picture, for example, accompanies the substantial section in Chapter 5 on the post-Impressionists.) Despite the lapses into odd terminology and obscurity, The Languages of Landscape offers many critical insights. The suggestion (p. 214) that a good deal of twentieth-century art does not so much depict the natural world as provide a substitute for its `sensual and affective appeal' is illuminating, as are the final reflections (pp. 237-238)--inspired by Monet's Giverny paintings--on landscape's `mediation' between the `basic alternatives of subjection to concentrated scrutiny and freely extravagant indulgence'.
The interest of readers of this journal will be aroused by the cover blurb's reference to the book's `new approach to understanding Western landscape art' through the use of `new methodologies deriving from sociology, anthropology, the study of rhetorical theory, and especially a version of visual semiotics'. I was unable, however, to detect any use being made of `new methodologies' in the social sciences, and while appeal is made to `rhetorical theory' and `visual semiotics', this is far less obtrusive than the blurb gives one to expect. The author himself is more accurate when informing his readers that `theory' will be used `only lightly and as a supporting framework' (p. 9). In fact, it is often used so lightly that one tends to forget the `framework' which officially supports this or that chapter or section. The discussion of the `naturalism' of Constable and others in Chapter 4, for example, proceeds in almost complete independence from the notion of `metonymy' which, at the outset, is said to inform it.
Lightly used or not, Roskill does advance a general theoretical position, one indicated by the title of his book. The subject of his study, he states, is not so much the appeal of landscape art as `the different kinds of "language" ... entailed in the creation and apprehension of it' (p. 8). And towards the end, it is claimed that `the most important aspects of verbal construal ... are all ones that can be transposed, as operative principles, to the recognition and reading of landscape imagery' (p. 229). In short, it is important to construct a `visual semiotics' which closely parallels, indeed apes, verbal semiotics. In keeping with this ambition, there are many references in the notes to well-known proposals in theoretical linguistics and the philosophy of language--Sperber and Wilson's `principle of relevance', for instance, or J. L. Austin's constative/performative distinction--together with suggestions as to their applicability to the understanding of paintings.
In my judgement, it is not only possible, but desirable, to read this book without being overly distracted by these gestures towards `visual semiotics', for most of the alleged similarities between understanding verbal utterances and `reading' paintings are strained and unconvincing--rather as, to go back a few years, were the similarities between language and the `codes' of, say, eating or dress alleged by structuralist theorists like Leach and Levy-Strauss. For example, there is surely no more than a vague analogy between the structural composition of paintings and that of sentences, one which is too stretched to warrant any serious reference to `a repertoire of landscape elements and a syntax for their arrangement' (p. 38).
It is inevitable that, in order to preserve the appearance of a `visual semiotics' parallel to a verbal one, Roskill is forced to employ key terms in an extended way. Thus, `Pragmatics' loses the relatively precise sense it has for theoretical linguists and seems to refer, for Roskill, to the study of just about anything that `links production to conditions of response', verbal or visual, including whole cultural backgrounds. It is, however, the familiar terms of rhetoric--in a chapter unduly inspired, one suspects, by Hayden White--which suffer most. If, for example, I understand Roskill's contorted definition of `synecdoche', it means for him the emphasizing by a painter of an essential aspect of his subject so as to infuse it with `an aura of persuasiveness'. This is a far cry from the verbal figure of synecdoche where, say, one refers to something by using a term normally employed to refer to a part of it. One can, perhaps, have something that deserves to be called visual synecdoche, as when a cartoonist draws a lock of hair and a moustache to represent Hitler, but it is nothing like that which Roskill has in mind.
There is, I believe, a general reason why there is little mileage in the idea of `visual semiotics'. Roskill rightly implies (p. 230) that interpretation of a painting is only an inferential process akin to understanding an utterance if the work has a `communicative end'. It is indeed because utterances, fairly generally, do have a communicative end--most notably, the stating of a proposition--that a reasonably systematic syntax, semantics, and pragmatics of a language are possible. Lexical elements and rules for their combination, for example, can be identified precisely in virtue of their systematic contribution to propositional sense. The trouble is that, pace Roskill, very few landscape paintings have a `communicative end' of comparable determinateness, and because so few do there can be nothing seriously comparable, in the case of art, to identifying the elements and combinatorial rules of a language. Doubtless we may speak, as Roskill perceptively does only a few lines before his remark on `communicative ends', of Courbet's The Oak at Flagey `evoking' both `extreme stability ... and the liberty invested in outdoor nature to grow and expand over time'. But one has only to look at the photograph of this painting on the opposite page to appreciate the gulf between responding to this `evocation' and the inferential processing of a statement about oaktrees, stability, or liberty.
Students and appreciators of landscape art will gain much from The Languages of Landscape. What they will not gain, however, is any confidence in the viability of the premises which inspire that title.
DAVID E. COOPER
University of Durham